Corrosion data now defines Renaissance Fusion’s liquid metal walls

Built for the Skyfall Hot campaign, the rig’s Helmholtz coils generate the roughly 1 tesla field used to hold the liquid metal layer against the chamber wall

(Image courtesy of Renaissance Fusion)

A new peer-reviewed paper from Renaissance Fusion, Sapienza University of Rome and Eindhoven University of Technology puts design-stage numbers behind flowing liquid metal plasma-facing components. Published in Nuclear Materials and Energy, it details material losses for three candidate steels exposed to liquid tin at 850°C, plus the current thresholds a test rig needs to hold a flowing metal layer against its chamber wall.

Corrosion testing narrows Renaissance Fusion’s steel choice

The team ran 24-hour static corrosion tests on three commercially available steels, 316L, 316Ti and 317, immersed in liquid tin at 850°C under argon. The 317 sample corroded through completely, with tin penetrating to the core from both sides. 316Ti lost 4mm of material and 316L lost 2mm. Line-scan analysis showed tin penetrating deep into each sample, followed by the formation of stannite compounds.

Researchers selected 316L as the structural material for the Skyfall Hot test rig based on this corrosion resistance combined with acceptable yield strength margins at temperature. Yield strength data existed up to 850°C for all three candidates, but only 317 had data above that threshold, which limited confidence in extrapolating the other steels’ properties higher.

Current thresholds for liquid metal wall attachment

Skyfall Hot is the current stage in Renaissance Fusion’s roadmap toward a fusion-relevant liquid metal blanket, sitting after the completed room-temperature Skyfall 1b demonstration and ahead of a planned helical-field experiment using pure lithium. The rig is designed to circulate liquid tin at 850°C inside a magnetic field of roughly 1 tesla, using injected current to push the flowing metal against the chamber walls through Lorentz forces.

OpenFOAM simulations run during the design process traced how that force builds with current. At 0.5 kilo-amperes the effect on the flow was negligible. At 1 kilo-ampere it was noticeable but still not enough to hold the streams against the walls. Above 1.5 kilo-amperes, the simulated cases reliably reached the target configuration of two stable, wall-attached streams. A separate analytical estimate, balancing Lorentz and centrifugal forces, put the minimum current at roughly 0.7 kilo-amperes, lower than the simulated threshold, so the physical rig carries extra current margin above that estimate.

The simulations also modelled what happens as the steel structure corrodes over the rig’s lifetime. A 25 percent reduction in wall thickness was found to reduce liquid layer thickness by 24 percent and increase flow velocity by 10 percent. More counterintuitively, the same modelling found that as the structure loses material, a larger share of the injected current reaches the fluid rather than short-circuiting through the solid, so wear improves electromagnetic control of the flow in principle, even as it thins the layer. That result comes from simulation at a specific operating point, not from measurement on the corroding rig itself.

Pebble trajectories and surface stability, still open

Renaissance Fusion’s blanket concept also requires embedding solid lead pebbles near the free surface as a neutron multiplier, targeting a tritium breeding ratio above 1.15 at a particle volume fraction of 19 percent. The paper introduces a modelling approach based on electromagnetophoresis, a force that acts on particles with different electrical conductivity than the surrounding fluid, and finds that particle size and injected current both work as levers for tailoring pebble trajectories, at least in the simplified geometry studied so far.

Separately, researchers applied a linear stability analysis to the completed Skyfall 1b experiment. Macroscopic droplets ejected from the liquid surface are a concern for power-plant compatibility, and the analysis catches this indirectly, by detecting the onset of unstable modes rather than simulating droplet formation itself. It picked up both free-surface and wall-based unstable modes. Video review of the Skyfall 1b run showed the liquid surface was not perfectly uniform once attached to the walls, though flow continuity was never disrupted. Measurement capability at the time could not confirm whether the irregularities matched the predicted instability wavelengths or came from elsewhere, such as the jet striking the wall.

Skyfall Hot sets the bar for what’s next

The paper frames Renaissance Fusion’s liquid metal wall concept as still conceptual, rating it Technology Readiness Level 2 on a nine-level scale the authors adapted specifically for liquid metal plasma-facing components. The stated goal for the current campaign is TRL 6, which in the paper’s own scale means operating the prototype under nominal conditions with interactions between subsystems characterised and performance validated using appropriate diagnostics.

As of the paper’s publication in August 2026, the Skyfall Hot rig had been built and its campaign was under commissioning, with results expected in the short term. Three related papers are cited as in preparation or under review, covering the full Skyfall 1b results, a detailed pebble transport model, and the complete linear stability formulation. The corrosion and current-threshold results published here are the first of that sequence to clear peer review.

Stay ahead in the fusion revolution explore more breakthroughs from leading innovators in clean energy technology.